This study reviews the evolution of electricity price forecasting (EPF) methods. We focus on a bibliometric survey to assess the strengths, weaknesses, and potential challenges that EPF methods present. We also discuss the ongoing debate between electrical engineers and econometricians over the effectiveness of statistical time series models versus artificial intelligence algorithms in EPF. Considering the role of power market integration, we further propose a hybrid CNN-BiGRU-Attention-XGBoost model to predict the hourly day-ahead electricity prices in Nord Pool markets. We demonstrate that our model statistically outperforms benchmark models in overall and hourly precision throughout the day across most regions. Our study provides a review and advancement of solutions for future complex EPF tasks.